Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: Problem with mlogit estimation

From   Maarten buis <>
Subject   Re: st: Problem with mlogit estimation
Date   Tue, 24 Aug 2010 08:21:18 +0000 (GMT)

--- On Tue, 24/8/10, wrote:
> global EU "dumat dumit dumfr dumes dumde dumnl dumgr"
> global demovar "dumsex age_lt_34 age_35_44 age_45_54
> age_55_64 age_65_74 age_75pl mi1 mi2 mi3 ms1 ms
> 2 ms3 ms4 ms5" 
> global sociovar "hh_sz ed1 ed2 ed3 dr01 dr2 dr4 dr5 ty4554
> ws1_2 ws2_2 ws3_2 ws4_2 ws5_2 ty4554"
> global consvar "dr01 dr2 dr3 dr4 dr5 cronill kids"
> global ecovar "Q1_hh_net_inc_byc Q2_hh_net_inc_byc
> Q3_hh_net_inc_byc Q4_hh_net_inc_byc Q5_hh_net_inc
> _byc Q6_hh_net_inc_byc"
> global finstress "make_ends_meet_comp mortgage_outs
> int_serv_ratio debt_serv_ratio fhealth"
> mlogit make_ends_meet $demovar $sociovar $consvar $ecovar
> $finstress $EU [pw=hwgt] if age!=. & emp_status_basic_3!=. 
> & educ_status_2!=. & migration_2!=. & hh_net_income!=. 
> Variables are qualitative. 
> Unfornately the results is wierd, because stata cannot
> calculate anymore the Wald test, the constant
>  and other beta coefficients

That is pretty common when you add that many variables, so
you'll need to simplify your model. You can do that by 
removing some of your explanatory variables, or by assuming
that make_ends_meet is ordinal and use -ologit-, -oprobit-,
or -slogit-.

Hope this helps,

Maarten L. Buis
Institut fuer Soziologie
Universitaet Tuebingen
Wilhelmstrasse 36
72074 Tuebingen


*   For searches and help try:

© Copyright 1996–2017 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index